September 12, 2008

U.S. Demand for Source-Differentiated Shrimp

By Jones, Keithly Harvey, David J; Hahn, William; Muhammad, Andrew

Estimates of price and scale elasticities for U.S. consumed shrimp are derived using aggregate shrimp data differentiated by source country. Own-price elasticities for all countries had the expected negative signs, were statistically significant, and inelastic. The scale elasticities for all countries were positive and statistically significant at the 1% level with only the United States and Ecuador having scale elasticities of less than one. For the most part, the compensated demand effects showed that most of the cross-price effects were positive. Our results also suggest that despite the countervailing duties imposed by the United States, shrimp demand was fairly stable. Key Words: CBS, conditional demand, countervailing duty, imports, scale elasticity, shrimp

JEL Classifications: C32, D12, Q13, Q22

This article addresses U.S. demand for sourcedifferentiated shrimp. There is a large body of literature on demand for different seafood species (Asche 1996; Asche, Salvanes, and Steen; Burton; Eales, Durham, and Wessells; Keithly and Diagne; Park, Thurman, and Easley). There is also a large body of literature on trade policy and conflicts in the seafood market (Asche 1996, 1997, 2001; Asche, Bremnes, and Wessells; Kinnucan; Kinnucan and Myrland 2002, 2006). However, the lack of studies on an important species like shrimp is an odd exception, since, according to the National Oceanic and Atmospheric Administration (NOAA), shrimp accounted for more than a quarter of the United States per capita seafood consumption in 2005.

For the last decade (1996-2005), U.S. shrimp imports have increased nearly 11% per year on average and have gained a greater share of total U.S. supply. At the same time, shrimp prices saw a more than 75% decline (see Figure 1). In 1996, domestic shrimp production and shrimp imports were 21% and 79% of total U.S. supply, respectively. Currently, imports account for over 90% of total U.S. shrimp supply (NOAA). In addition to significantly increasing, U.S. shrimp imports have become more concentrated in a few supplying countries. In 2004, six shrimp exporting countries supplied more than 70% of total U.S. imports, which were in excess of 1 billion pounds. These countries included: Brazil, China, Ecuador, India, Thailand, and Vietnam. Other major exporters of shrimp to the United States included Mexico, Bangladesh, and Indonesia.

Figure 1. U.S. Shrimp Imports, 1995-2006

In response to the increase in shrimp imports and falling shrimp prices, the U.S. shrimp industry filed petitions requesting that the U.S. International Trade Commission (USITC) and the U.S. Department of Commerce (USDOC) investigate whether shrimp imports were being sold to the United States below fair market value, and to also determine if exporters were receiving government support deemed injurious to the U.S. shrimp industry. In 2005, the USITC unanimously found that the U.S. shrimp industry had been injured by illegal dumping on the part of the aforementioned exporters (Brazil, Ecuador, India, Thailand, Vietnam, and China) and that a countervailing duty could be imposed on specific warm water shrimp imports from the six countries (Long; Zwaniecki).

The question then is would the countervailing duties imposed by the United States have any significant impact on shrimp import demand? Thomas and Ulubasoglu found that trade policies resulted in structural breaks in import demand for some commodities due to the change in access to substitutes. Mwega also found that trade policies that are designed to increase export earnings are more likely to have a large impact on import volumes.

In this study, we estimate U.S. demand for domestic and imported shrimp differentiated by exporting country using a Netherlands Central Bureau Statistic (CBS) demand system model. Parameter estimates are used to obtain elasticities of demand. We also explore the stability of shrimp demand in light of the countervailing duties imposed on some imported shrimp. Specific objectives are the following: (1) to econometrically estimate U.S. demand for shrimp (domestic and imported) where imports are differentiated by country of origin; (2) to estimate demand elasticities from the estimated demand parameters; (3) to test for monthly seasonality and stability of demand from each country.


In 2005 total U.S. landings of commercial shrimp were 261.1 million pounds and valued at $406.5 million. This was equivalent to 162.4 million pounds on a head-off weight basis. The Gulf States accounted for 88% of total landings that year while the South Atlantic and Pacific accounted for 7.4% and 4.4%, respectively. In 2004, U.S. production was 16% lower when compared to 2003. Since 2000. shrimp production had decreased 26%. During the last decade (1996-2005), U.S. commercial landings of shrimp (head-off basis) have averaged 190 million pounds. Above-average years included 2000 and 2001. where commercial landings were 218.5 and 201.0 million pounds, and below-average years included 1997 and 1998, where commercial landings were 179.0 and 173.3 million pounds, respectively. The year 2005 marked a record low for the decade. While U.S. shrimp production has declined by 56 million pounds since 2000, U.S. shrimp imports have increased to 467 million pounds. From 1996 to 2005, shrimp imports grew 107%. In 1996, U.S. shrimp imports were 720.9 million pounds, and in 2005, imports were 1.5 billion pounds. With the exception of 2005, imports have increased every year since 1996 (NOAA).

The increase in U.S. shrimp imports has been sustained by increases in United States per capita shrimp consumption. In 1980, per capita shrimp consumption was only 1.4 pounds. From 1990 to 1999, per capita consumption averaged 2.3 pounds, and in per capita consumption reached a record high of 4.2 in 2004. In 2005, per capita consumption was 4.1 pounds (NOAA). The rise in per capita consumption has made shrimp the top seafood product among U.S. consumers every year since 2001, surpassing can tuna and salmon, where per capita consumption in 2005 was 3.1 and 2.4 pounds, respectively (Johnson).

Under U.S. trade law, the administration of antidumping investigations is divided between the USDOC and USITC, where the USDOC determines whether imports subject to investigation have been sold in the United States at less than fair market value (Sharp and Zantow) and the USITC determines import injury to U.S. industries in antidumping, countervailing duty, and global safeguard investigations. If it is determined that exporters have injured U.S. producers, a countervailing duty is imposed on imports deemed by the USDOC as benefiting from subsidies generated by a foreign government or by a firm or person in that country (USITC). Based on the USITC determination, the U.S. Department of Commerce can issue a countervailing duty, which is enforced by the U.S. Customs Service. This is not necessarily imposed on all U.S. imports from a given country but on exporting companies as determined by USDOC to have benefited from different levels of government support (USDOC).

In December 2003, an association of U.S. shrimp farmers in eight southern states filed an antidumping complaint with the USITC against six countries that exported shrimp to the United States.1 The petitioned charged that exports from Brazil, China, Ecuador, India, Thailand, and Vietnam were dumped onto the U.S. market causing material damage to the domestic shrimp industry (Baughman; Bhattarcharyya; Blauer). This was not the first petition filed by the U.S. shrimp industry claiming injury from imports. In 1971, the National Shrimp Congress filled a petition with the USITC for import relief. A petition was also filed in 1984. In both instances the USITC ruled that U.S. shrimp farmers were not unfairly injured by shrimp imports (Diop, Harrison, and Keithly).

In the most recent petition filed, the primary points of contention were that the six named countries accounted for 74% of all shrimp imports and that imports from these countries increased from 466 million pounds in 2000 to 650 million pounds in 2002. They also charged that import prices in the targeted countries dropped 28% in the three years prior to the petition and that U.S. dockside prices fell from $6.08 to $3.30 per pound during that same period. Lastly, petitioners charged that higher tariff rates and sanitary requirements in other large importing countries made the U.S. market a dumping ground for shrimp exports that were rejected in markets such as the European Union (Bhattarcharyya).

In January 2005, the USITC ruled in favor of the U.S. shrimp industry, indicating that there was reasonable indication that the industry was materially injured by shrimp imports sold below market value. In a unanimous decision the USITC ruled that noncanned, mostly frozen shrimp and prawn imports from all six countries had hurt or threatened the U.S. shrimp industry (Long; Zwanicki;). The USITC ruling led to the imposition of antidumping duties where margins ranged from O to a high of 112.8%. The trade weighted duties for each country were 36.91% for Brazil, 49.09% for China, 7.30% for Ecuador, 14.20% for India, 6.39% for Thailand, and 16.01% for Vietnam (Bhattarcharyya).

Table 1. Antidumping Duties for Frozen Warm Water Shrimp Products

Company-specific duties, highest and lowest margins, and overall margins for each exporter are presented in Table 1. Duties were imposed on selected warm-water shrimp and prawns, whether frozen, wild caught (ocean harvested), or farm-raised (produced by aquaculture); head-on or head-off; shell-on or peeled; tail-on or tail-off; de-veined, cooked, or raw; or otherwise processed in frozen form. Frozen warm-water shrimp and prawn products included in the scope of the countervails, regardless of definitions in the Harmonized Tariff Schedule of the United States (HTSUS), are products that are processed from warm-water shrimp, frozen, and sold in any count size (Fact sheet-International Trade Administration [ITA], Department of Commerce). Data

The data consist of U.S shrimp production and prices and U.S. shrimp import data from eight importing countries for sixteen 10- digit HS codes. Six of these countries had antidumping duties imposed. Mexico and the rest of the world (ROW) had no antidumping duties applied. Brazil, though one of the countries levied with antidumping duties, was included with the ROW for two reasons. First, there were no reported imports from Brazil for several months during the study period. second, its share of imports by the United States was very small for the reported months. Though the countervailing duties varied among companies within the affected countries, data were not available for each company's imports. As such, import data for each of the Harmonized System (HS) codes was aggregated to a monthly total for each country and analysis of the countervailing duty impact was based on the countrywide rate of duty, rather than by company. Table 2 shows the descriptive statistics on U.S. consumption of shrimp by source country, 1995 to 2005.

The sixteen 10-digit HS codes could be divided into three broad categories-frozen, semiprocessed (brine, dry, and salted), and processed (prepared frozen, breaded, and canned). Frozen shrimp and prawn averaged 78% of all shrimp imported between 2000 and 2006, but for the period of study frozen shrimp imports have been steadily declining, from as much as 93% in 1990 to 71% in 2006 with most of the decline occurring in peeled shrimp. The frozen shrimp and prawn with shell-on of all count size ranged from 70-75% of the total frozen shrimp imported between 2000 and 2006 while peeled shrimp made up 25-30% of the total frozen shrimp imported. Semiprocessed shrimp, which consists of dry, salted, or brined shrimp, made up 1% of shrimp imports in 2006 and has steadily declined from around 3% in 1990. Processed shrimp (prepared frozen, breaded, and canned) has seen the largest increase in the imports over the study period.

Table 2. Descriptive Statistics on U.S. Consumption of Shrimp by Source Country, January 1995 to December 2005

Monthly import quantities and expenditures from each country were obtained from the U.S. Department of Agriculture, Foreign Agricultural Service, and Foreign Agricultural Trade Statistics. All expenditures are on a free on board (FOB) basis, excluding transportation costs, insurance and custom duties, making import prices fairly representative of the wholesale U.S. domestic price. Using expenditures and quantities, per-unit values ($/lb) for each country were calculated.

Because of the way it is constructed, the endogenous variables of the CBS demand system sum to 0 in every time period, which makes the error terms sum to 0 as well. In order to estimate the system, we have to drop an equation. The estimates will be invariant to the equation dropped; we happened to drop the equation for U.S. shrimp.

CBS models are often estimated with an intercept. The intercept is generally justified as a taste-shift term; a non-0 intercept produces demand changes unrelated to changes in prices, expenditures, and/or scale. (In our derived-demand case, an intercept could measure some combination of taste and technology shifts.) We are using monthly data; we use 12 monthly dummies in each equation rather than an intercept.

The monthly dummies' coefficients must also be consistent with the budget constraint; this means that all the January coefficients have to sum to 0, and all of February's, and so on. We looked at three different hypotheses regarding the seasonal effects. It might be the case that there is no seasonality in derived shrimp demand. If so, we could have used an intercept rather than a set of monthly dummies. We compared models with only an intercept to one with a full set of dummies. As noted above, the monthly dummies represent taste/technology shifters. Shrimp demand might change month to month, but still be stable on an annual basis. If each shrimp source's monthly dummies sum to 0, (January plus February . . . ) then demand is stable on a year-to-year basis. We can combine the no- seasonality and stable-annual-demand models by estimating a model without either intercepts or dummies.

Empirical Results and Discussion

Since the dynamic versions of the model are nonlinear and the negativity conditions are nonlinear, inequality restrictions, the model was estimated using nonlinear maximum likelihood estimation. Sixteen versions of the model were estimated; each combination of the four dynamic and the four monthly dummy/intercept hypotheses. Table 3 shows the hypothesis tests based on the likelihood ratio test. Demand dynamics are consistent with an autoregressive process; the autoregression is statistically significant. Shrimp demand has a statistically significant seasonal pattern, but is stable on an annual basis.

Estimated conditional price and share demand coefficients are reported in Table 4. For the most part, the compensated" demand effects, the "C^sub ij^" coefficients, showed that most of the cross- price "C^sub ij^" terms are positive. Using these estimates and average budget shares for the sample period, own- and cross-price elasticities and scale/expenditure elasticities were generated. These estimates are presented in Table 5. The standard errors in parenthesis were generated using 1,000 bootstrap iterations. Green, Roke, and Hahn show that the bootstrap covariances are more accurate measures of the small sample covariances than the more conventional approaches (e.g., the Rao approach) that use asymptotic approximations.

The conditional own-price elasticities represent both the substitution and the income effect of price changes. These elasticities are conditional on total expenditure on shrimp. The own- price elasticities for all countries had the expected negative signs and were statistically significant and inelastic (less than -1), implying that it is possible for these countries to increase revenue by lowering quantities supplied. This also implies that an increase in these countries' shrimp prices would result in a less than proportionate decrease in the quantity of shrimp demanded from them by the United States. The least inelastic was Thailand, -0.825. A 1% increase in the Thailand shrimp price will result in a decrease in their quantity of shrimp demanded by only 0.825%. The most inelastic was the ROW. A 1% increase in the ROW shrimp price will result in a decrease in their quantity of shrimp demanded by only 0.242%. The paucity of studies on shrimp import demand limited the ability to compare results.

Table 3. Hypothesis tests for the CBS Model

Table 4. Estimated Price and Scale coefficients for the CBS Model, with Dynamic Adjustments Replaced with AR1

The scale elasticity measures the degree by which the amount of shrimp demanded from the United States and importing countries change when U.S. overall shrimp demand changes; this scale elasticity is also the elasticity of total wholesale shrimp expenditure. Scale elasticities for all countries were positive and statistically significant at the 1% level. The estimated scale elasticity for Mexico, India, Thailand, Vietnam, and China were greater than one. India had the largest scale elasticity of demand of 1.740, which implies that a 1% increase in overall U.S. shrimp demand would increase Indian shrimp import demand by 1.740%. Ecuador and the United States had scale elasticities of less than 1, though positive suggesting that a 1% increase in total U.S. shrimp demand would result in a less than 1% increase in shrimp demand from these countries.

Table 5. Estimated Price and Scale Elasticities for the CBS Model, with Monthly Seasonality but Stable Year-to-Year Demand and AR1

Table 6. Estimated Monthly Dummies with Seasonality but Stable Year-to-Year Demand and AR1

As noted above, the elasticities in Table 5 are conditional on total shrimp expenditure. These cross-price elasticities account for both the substitution effects and expenditure effects of price changes. In a CBS specification, the "pure" substitution effects are determined by Cjj terms. Most of these cross-price terms are positive. However, the cross-price elasticities were very low, with 0.086 being the largest in absolute value. For the most part, signs on the cross elasticities were negative. This demonstrates that the expenditure effects of price declines tend to outweigh the pure substitution effects. For example, China's cross elasticity were negative with all other countries. A few cross elasticities were positive; for example, a 1% increase in the price of U.S. shrimp would result in a statistically significant but small increase in imports in shrimp from Thailand of 0.031%.

Stability of Demand

The premise associated with the antidumping procedure is that it has a measurable impact on import behavior and that dumping margins have a negative impact on dumped imports with stronger reactions for larger margins. The expectation, then, is that countervailing duties will destabilize the demand for shrimp imported from countries affected by the antidumping duties. The results shown in Table 6 demonstrate that although a high degree of monthly seasonality existed shrimp import demand form all of the countries, the overall demand was fairly stable. Conclusions

The aim of the paper was threefold. First, we empirically estimated the total U.S. demand for shrimp and the conditional import demand for shrimp consumed in the United States with an econometric model. From this we calculated conditional import demand elasticities from estimated demand parameters. Finally, we tested for seasonality and stability of shrimp import demand in the face of a countervailing duty imposed on imports from certain countries.

The own-price elasticities for all countries had the expected negative signs, were statistically significant, and inelastic. The scale elasticities for all countries were positive and statistically significant at the 1% level with only the United States and Ecuador having scale elasticities of less than 1. For the most part, cross elasticities were negative, implying that shrimp demand exhibited a complementary relationship between countries. In a few cases, cross elasticities were positive, suggesting a substitute relationship.

Our results from this study suggest that despite the countervailing duties imposed by the United States on six major shrimp exporting countries, shrimp demand from these countries was fairly stable although there was a high degree of monthly seasonality. The monthly seasonality was expected since production for most countries is seasonal.

[Received May 2007; Accepted December 2007.]

1 The association of shrimp farmers in the South is often referred to as the Southern Shrimp Alliance or the Ad Hoc Shrimp Trade Action Committee.


Anderson, G.J., and R.W. Blundell. "Estimation and Hypothesis Testing in Dynamic Singular Equation Systems." Econometrica 50(1982): 1559-1571.

Asche, F. "A System Approach to the Demand for Salmon in the European Union." Applied Economics 28(1996):97-101.

_____. "Trade Disputes and Productivity Gains: The Curse of Farmed Salmon Production." Marine Resource Economics 12(1997):67- 73.

_____. "Testing the Effects of Anti-dumping Duty: The U.S. Salmon Market." Empirical Economics 26(2001):343-55.

Asche, F., K.H. Bremnes, and C. Wessells. "Product Aggregation, Market Integration, and Relationships Between Prices: An Application to World Salmon Market." American Journal of Agricultural Economics 81(1999): 68-81.

Asche, F., K.G. Salvanes, and F. Steen. "Market Delineation and Demand Structure." American Journal of Agricultural Economics 79(1997): 139-50.

Barton, A.P. "Maximum Likelihood Estimation of a Complete Demand System of Equations." European Economic Review l(1969):7-73.

Baughman, L.M. "Shrimp Antidumping Petition Would Jack Up Prices to Shrimp-Consuming Industries." The Trade Partnership, Washington, DC. 2004. Internet site: www.tradepartnership. com (Accessed April 23, 2007).

Bhattarcharyya, B. "The Indian Shrimp Industry Organizes to Fight the Threat of AntiDumping Action." WTO, Managing the Challenges of WTO Participation: case Study- case Study 17. Geneva, Switzerland. 2005. Internet site: res_e/booksp_e/ casestudies_e/ case 17_e.htm (Accessed April 23, 2007).

Blauer, R. "Shrimp Imports Increase Despite Confirmed Antidumping." Fish and Seafood Products: Market News, March 2007. Foreign Agricultural Service, USDA, Washington, DC. 2007. Internet site: Newsroom/Articles/shrimp_imports.asp (Accessed April 23, 2007).

Burton, M.P. "The Demand for Wet Fish in Great Britain." Marine Resource Economics 7(1992): 57-66.

Deaton, A.S., and J. Muellbauer. "An Almost Ideal Demand System." American Economic Review 70(1980a):312-26.

_____. Economics Consumer Behavior. New York: Cambridge University Press, 198Ob.

Diop, H., R.W. Harrison, and W.R. Keithly, Jr. "Impact of Increasing Imports on the United States Southeastern Region Shrimp Processing Industry 1973-1996." Selected Paper, American Agricultural Economics Association, Nashville, TN, 1999. Internet site: http://agecon.lib.umn. edu/cgi-bin/ 1460&ftype=. pdf (Accessed December 15, 2006).

Bales, J., C. Durham, and C.R. Wessells. "Generalized Models of Japanese Demand for Fish." American Journal of Agricultural Economies 79(1997):1153-1163.

Green, R., D. Roke, and W. Hahn. "Standard Errors for Elasticities: A comparison of Bootstrap and Asymptotic Standard Errors." Journal of Business Economics Statistics 5(1987): 145-50.

Johnson, H. "Top 10 U.S. Consumption by Species Chart." 2007. Internet site: www.aboutseafood. com/media/top_10.cfm (Accessed April 23, 2007).

Keithly, W.R., Jr., and A. Diagne. "An International Import Demand and Export Supply Model for Shrimp and Impacts of Changes in World Production on the U.S. Dockside Price." International Institute of Fisheries Economics and Trade Conference Proceedings, 1998.

Keller, W.J., and J. Van Driel. "Differential Consumer Demand Systems." European Economic Review 27(1985):375-90.

Kinnucan, H.W. "Futility of Targeted Fish Tariffs and an Alternative." Marine Resource Economics 18(2003):211-24.

Kinnucan, H.W., and O. Myrland. "The Relative Impact of the Norway-EU Salmon Agreement: A Midterm Assessment." Journal of Agricultural Economics 5 3(2002): 195-219.

_____. "The Effectiveness of Antidumping Measures: Some Evidence for Farmed Atlantic Salmon." Journal of Agricultural Economics 57(2006):459-77.

Long, D. "U.S. Shrimp Industry Wins Final Antidumping cases Against Six Countries." Southern Shrimp Alliance Press Release. 2005.

Mwega, F. "Import Demand Elasticities and Stability during Trade Liberalization: A case Study of Kenya." Journal of African Economies 2(1993):381-416.

NOAA (National Oceanic and Atmospheric Administration, U.S. Department of Commerce). "Fishery Statistics of the United States, 2005." Silver Springs, MD. 2007. Internet site: stl/fus/fus05/index.html (Accessed April 23, 2007).

Park, H., W.N. Thurman, and J.E. Easley, Jr. "Modeling Inverse Demand for Fish: Empirical Welfare Measurements in the Gulf and South Atlantic Fisheries." Marine Resource Economics 19(2004):333- 51.

Sharp, D., and K. Zantow. "Attribution of Injury in the Shrimp Antidumping case: A Simultaneous Equations Approach." Economics Bulletin 6(2005):1-10.

Theil, H. "Applied Economic Forecasts." North Holland, Amsterdam. 1966.

Thomas, D.D., and M.A. Ulubasoglu. "The Impact of Trade Liberalization on Import Demand." Journal of Economie and Social Research 4(2004): 1-26.

U. S. Department of Commerce. International Trade Administration. Federal Register. Vol. 63, No. 227, November 1998.

U.S. International Trade Commission. "Certain Frozen or Canned Warm Water Shrimp and Prawn from Brazil, China, Ecuador, India, Thailand, and Vietnam." Investigation Nos. 731-TA-1063-1068 (Final). Publication # 3748, January, 2005.

Zwaniecki, A. "U.S. Duties Imposed on Frozen Shrimp from Six Countries Trade Panel to Review India, Thailand cases to Assess Tsunami Impact." The Washington File, January 2005. Bureau of International Information Programs, U.S. Department of State, Washington, DC. Internet site: display.html?p=washfile0 -english&y=2005&m= January &x = 20050106172826SAikceinawz 0.3572199&t=x0 archives/x0 architem.html (Accessed April 23, 2007).

Keithly Jones, David J. Harvey, and William Hahn are economists with the Animal Products, Grains, and oilseeds Branch, Markets and Trade Economics Division, Economic Research Service, USDA, and Andrew Muhammad is assistant professor, Department of Agricultural Economics, Mississippi State University.

The views expressed here are those of the authors, and may not be attributed to the Economic Research Service or the U.S. Department of Agriculture or Mississippi State University.

Copyright Southern Agricultural Economics Association Aug 2008

(c) 2008 Journal of Agricultural and Applied Economics. Provided by ProQuest LLC. All rights Reserved.